Ino Get Model Config

Every model has a "right" way to run it - the sampler, the scheduler, the steps, the guidance, which text encoders and VAE it wants. Remembering that for one model is easy; juggling five is how workflows drift into "it used to look better." Ino Get Model Config is the Ino Nodes pack's answer: a dropdown of model presets that returns the entire configuration - sampling settings and component download configs - as one JSON string, so the rest of the graph runs on a single source of truth.

It's part of the Ino Nodes pack from nobandegani, and it's the front door to the pack's config-driven workflow story. The presets live in data/models.json, which ships with the pack and holds real entries like the Flux 1 dev configuration with its euler/simple 30-step settings, guidance 3.5, and the exact HuggingFace repos for its UNet, CLIP1, CLIP2 and VAE.

What you get

Two inputs, both dropdowns, both populated from the pack's models JSON at load time:

• model_name - "The name of the Model," per the author's tooltip. This is the one you'll set: pick the preset.
• model_id - "The id of the Model." A numeric alternative; setting either one selects the config.

Pick one and the node hands back:

• model_config - the full preset as a JSON string. The important one.
• model_name / model_id - the resolved name and numeric id.
• success / message - reporting.
• datetime_iso - a UTC timestamp of when the config was read, which is handy for logging and version-tracking which config state a run used.

That model_config string is the workhorse. Feed it into Ino Get Sampler Config and the graph builds a guider, sampler, and sigmas from it. Feed it into Ino Get Model Download Config and it splits into the per-component download configs for UNet, CLIP1, CLIP2 and VAE. One string, two consumers - that's the whole design.

Where it fits

If you run the same couple of models repeatedly, this node turns "remember the right settings" into "pick from a list." It also makes workflows shareable without losing the recipe: the config carries the sampler settings (e.g. euler/simple, 30 steps, guidance 3.5, cfg 1 for Flux) so a downloaded workflow isn't silently running with whoever's stale defaults.

The honest caveat: the preset list is whatever the pack ships, and it's Flux-centric by default. A model you use that isn't in models.json won't be in the dropdown - the pack has config-editing nodes (Ino Update Model Config, Ino Create Model Config) for building your own, but out of the box you're limited to the shipped set.

Installing it

It ships with the whole Ino Nodes pack. ComfyUI Manager: search "ComfyUI Ino Nodes", install, restart. Or manually:

cd ComfyUI/custom_nodes
git clone https://github.com/nobandegani/ComfyUI-InoNodes.git
cd comfyui_ino_nodes
pip install -r requirements.txt

Restart after. Pack-wide caveats apply: it's built on ComfyUI's V3 schema (README wants v0.18.1+), and requirements.txt installs inopyutils, the helper library every node imports.

Common issues

If the dropdowns look wrong or empty, the JSON was loaded when the pack loaded - update the pack or edit models.json and restart ComfyUI before expecting changes. And note that setting both name and id is fine, but the name takes priority in the lookup, so leave the other on its default if you want no surprises. Search "Ino Get Model Config" in the node menu (grouped under InoSamplerHelper).